DMOS devices are widely used in circuits. Typically, for low-side DMOS, because the source potential may be at reference zero, both the source and body in termination are connected to an isolation ring. So no additional layer is needed to execute isolation with other devices or to improve the breakdown voltage. As a result, the termination area is quite small. However, for high-side DMOS, due to the high-side DMOS's work condition (e.g., the high voltage potential across the high-side DMOS and the source potential is float), full isolation is required). A common way to achieve full isolation is to extend the Nwell region, as shown in FIG. 1A, so that the source and body of the DMOS are enclosed by the Nwell region. However, this may cause low breakdown voltage between the Pbase region and the Nwell region in termination.
An improved way is to insert a Pwell region in the Nwell region in termination. As shown in FIG. 1B, the Pwell region 11 which partially surrounds the Pbase region 12 is embedded in the Nwell region 10. The Pbase region 12 acts as the body region. This approach changes Pbase-Nwell breakdown (the breakdown voltage between the Pbase region 12 and Nwell region 10) to Pwell-Nwell breakdown (the breakdown voltage between the Pwell region 11 and Nwell region 10). Generally, the Pwell-Nwell breakdown voltage is much higher than Pbase-Nwell breakdown voltage in DMOS. However, the inserted Pwell requires a large area and increases the size of the device.